The proposed dissertation will examine the relative effectiveness of social support provided by health and mental health services and that provided by "natural" support systems on the health of a mentally ill, homeless population. Specifically, it will compare the effect of perceived social support provided by programs and perceived social support provided by kin, friends and other primary group members on physical health, psychiatric symptomatology, and physical and psychosocial functioning. Data will be obtained from an NIMH-supported outcome study of a sample of 97 mentally ill homeless individuals served by one of five nontraditional programs evaluated by the Department of Epidemiology of Mental Disorders Research Department at New York State Psychiatric Institute. Services received by the subjects include health care, mental health care, substance abuse treatment, shelter, income maintenance, and advocacy, and are provided directly or by referral. Data on demographic variables, residence history, tratment history and psychiatric diagnosis were obtained with standardized instruments and a questionnaire developed for the NIMH-supported project. The outcome measures, self-reported health, severity and type of psychitric symptoms, physical functioning, and psycosocial functioning, were taken at entry into the study (PRE) and six months later (POST).m Perceived social support was also measured at PRE and POSt levels, using an instrument developed for this dissertation project. Life events were assessed for a small subsample (N=50) using a standardized schedule modified for the dissertation project. The proposed study will use a quasi-experimental design to control for baseline health and demographic variables and to determine the effect of social support and services rendered on health status change over six months. It will also test three patterns of program and informal social support use: Program Dependency, Manipulation of Resources, and Social Margin. Finally, it will attempt to validate dimensions of support. The proposed dissertation addresses delivery of services to an economically disadvantaged, hard-to-serve population. In particular, it examines whether or not "natural" supports can be used to reduce the need for formal, more expensive services. It has practical implications for all medical and mental health services which have a dual function: treatment and (often unintended) social support. It will also contribute to knowledge about how to serve an increasingly needy population -the homeless mentally ill.